Germany Seeks The Most Secure IT In The World

It could be all over for Wintel in Germany if a proposed law on security in IT bears fruit.“The Federal Government wants Germany’s IT systems and digital infrastructure to be the most secure in the world” The idea is that every IT system must be as secure as possible whatever it takes proportionate to the risks. That would imply GNU/Linux or UNIX operating systems should be used everywhere. That cuts out M$ and possibly Apple. M$’s software is so utterly complex that I doubt anyone could with a straight face pronounce it secure and Apple’s is not even in the running.

Combine that with the already heavy use of GNU/Linux in the country and I think it’s a near certainty GNU/linux will be blessed. It’s more than operating systems though. Will Java make the cut? How about Adobe’s stuff, eh? and Flash? I wonder if Germany will evolve to a “certification” process for software? That could cut out a lot of non-Free software right there. How can you certify binary-only software? How can you certify binary-only updates made at noon on a work-day?

This will be something to watch no matter how it turns out. Practically, I think it is most likely that Germany will ship a government distro with only well-tested software in the vault. I recommend they start with Debian GNU/Linux.

See Germany unveils draft Federal IT security act.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology and tagged , , , , , . Bookmark the permalink.

35 Responses to Germany Seeks The Most Secure IT In The World

  1. TEG rudely wrote, “did I mention you actually wanted the statistics by each specific country or place rather than the world as a whole?”

    I have the shares in each country thanks to StatCounter but I have no way of proving that their “Worldwide” number means anything from the documentation they provide. It sounds like they are just adding up all the page-views globally and sorting them by OS. That’s not the right way to do it if we want the average page-view share by GNU/Linux over the whole world which is what people are talking about when they say “~1%”. There is data on the web about how many people are connected but it’s not about legacy PCs but all forms of connections. I care most about the legacy PC and the share of them that run GNU/Linux. I don’t want a number that’s heavily biased by data from USA warping the number for the world. To do that we need proper weighting: connected legacy PCs per country, and page-views (not share of page-views) per country. StatCounter has only revealed how many websites are being counted per country. That may or may not be relevant. Without that proper weighting, StatCounter’s worldwide number is meaningless.
    e.g. $latex N = \sum n_{i}$ where $latex n_{i}$ is the number of connected legacy
    PCs in the $latex i^{th}$ country.

    $latex w_{i} = \frac{n_{i}}{N}$

    $latex W = \sum w_{i} = 1$

    World Wide Share $latex = \sum w_{i}Share_{i}$

  2. That Exploit Guy says:

    I don’t always want “share in the good old USA”. If I want the global share I need a way to combine the individual shares in a linear combination that represents an average, you know the value that if everywhere the same would give the same number of page-views?

    What do the page views in USA have to do with the page views in Ethiopia? Nothing. That’s what. You are talking about an approach that only (kind of) makes sense if you are looking at the market share by each OS in the world as a whole, not when, say, you want the market share by each OS in Turkey, India, the UK, Brazil, etc. etc. Even then, any “weight” you apply is at best arbitrary and inherently implies that some machines generate more page views than others. In other words, you are merely distorting the statistics to suit your interpretation.
    Oh, by the way, did I mention you actually wanted the statistics by each specific country or place rather than the world as a whole? Yeah, why are you changing the subject now? Is that because you have been wrong about NAT interfering with StatCounter numbers and the numbers turn out to be unfavourable to your expectations?
    Typical Robert Pogson.

  3. olderman wrote, “Samba is a hack beloved by freetards. It doesnt get used in a production environment otherwise.”

    I, for one, don’t love Samba. I avoid it like the Plague that M$ intended SMB to be. M$ foisted that on the world to make file-sharing “easy” but insecure and complex. They also made it a moving target so FLOSS could never do it right. That said, many storage devices have Samba built in and rely on SMB to shift files on networks. A lot of such devices do get used in production environments.

    Samba is one of the reasons I recommend organizations do a sudden migration instead of a gradual migration. It’s almost impossible to get different versions of that other OS to agree on SMB let alone GNU/Linux. The first time I ever saw “7” it refused to share a printer with XP. When I used Samba with M$’s server OS I did see 30s delays observed by WireShark. XP machines were taking 2minutes to boot and they saw it too.

    Further, enough businesses use Samba that other businesses support Samba:
    e.g.

    • Net Direct is Canada’s premier Linux solutions provider and an IBM premier partner. We offer a wide range of Linux solutions including Samba and Windows integration solutions. We also offer Samba consulting, support and system integration solutions.
    • Savoir Faire Linux: Une équipe exceptionnelle de 100 consultants en logiciels libres basés à Montréal, Québec, Ottawa et Paris qui, depuis 15 ans, fournit des services de consultation, d’intégration, de développement, de soutien technique et de formation sur les technologies ouvertes.[My rusty French translatioin:”An exceptional team of 100 consultants in Free Software based in Montreal, Quebec, Ottawa and Paris, who, for 15 years, have provided consultation, integration, development, technical support, and development of open technological services”]

    and then there’s RedHat, IBM and a bunch of others you may have heard of. Someone’s using it in production.

    I find it interesting that RedHat recommends using NFS rather than Samba between GNU/Linux machines. That’s my belief. Use NFS and chuck that other OS. Make the world a better place.

    I found an article from IBM, The Samba performance challenge with audio. The bottom line is that they can do a crazy throughput with good old NFS and some SSD caching and networking magic. Why let Samba or SMB slow it down?

  4. TEG wrote, “use the drop-down menu and filter out page views that are from the USA. You don’t “weight” the page views.”

    I don’t always want “share in the good old USA”. If I want the global share I need a way to combine the individual shares in a linear combination that represents an average, you know the value that if everywhere the same would give the same number of page-views? To do that I have to use a weighted sum because USA and other countries are severely over-weighted. Their page-views happen much more than the other places. ie. I can’t just add up the shares because then GNU/Linux would get over 100% etc. Since they don’t publish the number of page-views, just the shares, there’s no rational way to do it. The random sampling method is as good as any. It’s called Montecarlo Integration by the way…

  5. olderman says:

    “ThePluralOfUnsubstantiatedBSisFact(TM)”

    Perhaps, Mr. Kitty.

    But Dougie doesn’t even throw as good a line of BS as Lil’ Hammie does.

    Then again, I wouldn’t expect him to be able to given his posts.

  6. olderman says:

    “Samba…. a hack…for freetards”… now we all know where you stand.

    Wherever I stand I’ll bet good money that its someplace that you aren’t qualified to even be cleaning the floors.
    .

    “Yes, fixing Win-Dohs problems is major reoccurring theme and it pays well.”

    Nope, my most recent problems involved assisting the senior linux sysadmin in troubleshooting some weird problems on a SAN attached Linux host.m It turned out to be a combination of firmware on the HBA and some fun and games with the linux multipath drivers. Fortunately Red Hat issues a fix for that.

    “But, since we are discussing secure IT, Windows sure doesn’t live up to that fact does it?”

    Can we say, Heartbeat, openssh and Bash?

    “So does this mean you won’t be rebooting your AD & DC servers during lunch-hour? I mean, 10-second boot-times are just awesome, whats the matter you scared?

    Nope just not as stupid and ignorant as you apparently are for suggesting it.

    “Your shop? LOL.. you don’t own a IT shop, nor do you oversee anyone.”

    How do you know Dougie?

  7. dougman says:

    “Samba…. a hack…for freetards”… now we all know where you stand.

    Yes, fixing Win-Dohs problems is major reoccurring theme and it pays well.

    But, since we are discussing secure IT, Windows sure doesn’t live up to that fact does it?

    So does this mean you won’t be rebooting your AD & DC servers during lunch-hour? I mean, 10-second boot-times are just awesome, whats the matter you scared?

    Your shop? LOL.. you don’t own a IT shop, nor do you oversee anyone.

  8. olderman says:

    Oh … and Dougie..

    Its oldERman, not OLDman.

    Not getting the details right is not good from someone in IT like yourself.

    But just keep using that Linux recovery disk and you should be OK.

  9. olderman says:

    “See, I knew OLDman would agree with me, Samba is a much better solution in the long run, he is most wise.”

    You are a comedian dougie. Lets cut the bullshizze shall we. Samba is a hack beloved by freetards. It doesnt get used in a production environment otherwise.

    But then again dougie you wouldn’t know a production environment spending so much time fixing PC’s with your trusty Linux recovery disk.

    “10-second boot times with Samba is awesome, I dare you to reboot your M$ AD & DC servers during working hours…LOL.”

    I suppose that supposed to annoy me, eh Dougie. SO Sad. Only someone ignorant of productions environments and their SLA’s would make such a statement.

    You would’nt last very long in my shop, dougie.

    Best to keep selling your snake oil to the unwary and leave the real work to the professionals.

    LOL.

  10. dougman says:

    See, I knew OLDman would agree with me, Samba is a much better solution in the long run, he is most wise.

    https://www.samba.org/samba/news/releases/4.0.0.html

    10-second boot times with Samba is awesome, I dare you to reboot your M$ AD & DC servers during working hours…LOL.

  11. olderman says:

    “Samba is a far better solution then Active Directory.”

    Suuure doougie, sure.

    Now why don’t you run along and rescue someones hard drive like a good little tech.

  12. That Exploit Guy says:

    OLDman – When a problem involves Linux, it’s always the admin’s fault. When it’s a proprietary solution, it’s bad software.

    ThePluralOfUnsubstantiatedBSisFact(TM)

  13. dougman says:

    Samba is a far better solution then Active Directory.

    OLDman – When a problem involves Linux, it’s always the admin’s fault. When it’s a proprietary solution, it’s bad software. You must be new here.

  14. That Exploit Guy says:

    Yes, but what weights to apply?

    What do weights have to do with anything? You asked:

    So what do those “global” stats mean when a disproportionate share of the page-views are from USA?

    The answer is to simply use the drop-down menu and filter out page views that are from the USA. You don’t “weight” the page views. There is no valid justification for the “weights” you apply to the page views. Instead, you should use the page views themselves to determine the share by each operating system, like this. Seriously, stop fixing what is not broken with your stupid, bungled math.

  15. olderman says:

    “I’ve used AD and it’s the pits. Slow and halting”

    So tell us Robert Pogson, how long ago was that? on Windows server 2003?

    Our production Active Directory instance running windows Server 2008 R2 services 10K+ active users and has 130K+ active accounts, and is NOT slow and halting. And that same instance spans the globe as well.

    Eh?

  16. TEG wrote, “Have you tried filtering the results with the “Region” drop-down menu?”

    Yes, but what weights to apply? StatCounter obviously has no idea and neither do I. You can’t weight the regional numbers with PC presence nor page-count presence nor regional population because none of those relate well or we do not know the numbers. If we just average the regional numbers we might come close to reality but we might not. The API to do that is a lot of clicks on their pages. If you want to waste a day and send me the numbers you’re welcome to do so…

    Here, let me help you. I wrote a little programme that randomly selects 21 items from a list of 238 domains covered by StatCounter:
    program random_select;
    var s:string[3];t:array[1..238] of string[3] ;f:text;
    j,k,n:longint;
    begin
    assign (f,'TLD');
    reset(f);n:=0;
    while not eof(f) do begin readln(f,s);inc(n); t[n]:=s end;
    reset(f);
    writeln('there are ',n:3,' top level domains in the list');
    for j:=0 to 20 do
    begin
    k:=random(n)+1;writeln(k:3,' ',t[k]);
    end;
    close(f)
    end.

    Running it gets this the first time through:
    ./random_select
    there are 238 top level domains in the list
    131 LY 0.46%
    142 MO 0.68%
    171 PH 0.59%
    201 ST 1.49%
    144 MQ 2.27%
    205 TC 1.35%
    130 LV 1.60%
    202 SV 1.55%
    101 IN 1.63%
    149 MV 0.61%
    154 NA
    92 HK 0.99%
    105 IT 2.18%
    71 FM 0.52%
    213 TN 0.86%
    14 AU 0.87%
    230 VN 0.83%
    65 ER 2.56%
    92 HK
    114 KM 0.84%
    189 SC 1.24%

    One of those is HK which occurs twice. We can throw out one. Another is NA which is a region larger than a country, so we can throw that out. That leaves 19 items to sample. The third column I typed in with data from StatCounter. The average is 1.22%. That’s just as valid as StatCounter’s global number of 1.30%. Anyway, aren’t you glad to know that the infectionprevalence in Ethiopia is spreading to Eritrea? 😉

  17. That Exploit Guy says:

    They state they are unbiased yet they remove “bots” without any mention of how they do that.

    Bots’ agent strings (or lack thereof) are usually pretty easily spotted. Unless a bot is deliberately masquerading itself by pretending to be a browser, then I don’t see why bots are even an issue to begin with. Are you suggesting that some sort of global conspiracy is going on where bots are being used to manipulate StatCounter statistics?

    So what do those “global” stats mean when a disproportionate share of the page-views are from USA?

    I dunno. Have you tried filtering the results with the “Region” drop-down menu?

  18. TEG wrote, “StatCounter states rather clearly that it uses page views rather than unique IP, so NAT is simply an non-issue for its global stats.”

    So what do those “global” stats mean when a disproportionate share of the page-views are from USA? Does a page-view from USA count as 12 times more than one from China or India? USA has only ~20% of the world’s PCs… China has more PCs than USA but gets many times fewer page-views counted. Turkey, a country of 77 million is counted way more often than Germany, a modern country with 80 million people. Is it just coincidence that Germany with 2.5% share is counted so little while Turkey with 0.44% GNU/Linux is counted so heavily? France, another heavy user of GNU/Linux is way down the list. I see many biases here. What is the mechanism? If it’s not NAT/firewalling its something just as vigorous.
    ” 4.2 billion – United States
    1.6 billion – Turkey
    1.2 billion – India
    776 million – United Kingdom
    759 million – Brazil
    537 million – Germany
    517 million – Canada
    367 million – Indonesia
    340 million – China
    336 million – Thailand

    We don’t apply artificial weights to our data (but our users can do so if they wish).”

    I’d say the counts per country is an artificial weight if there ever was one.

  19. Kevin Sorbo wrote, “That’s a meaningless blanket statement. It depends on the size of the school. Some deployments are quite small. How many systems in your school up in the Innukistan outback, for instance? “

    What does size matter? The principles are the same. People want reliable and fast IT. Most schools I was in had at least a computer lab and a PC on teachers’ desks. The largest had two labs, clusters in several places and a few available to each classroom, 136 GNU/Linux seats and 6 GNU/Linux servers. The largest system running that other OS was about the same size and was much more trouble to operate.

    Later, he pushes M$’s products, “GPO by itself increases operating efficiency by leaps and bounds. Even in a heterogenous environment, AD and GPO for the windows systems, Puppet or Chef for the *nix systems. You learn a thing or three on efficiency when you’re managing thousands of systems and tens of thousands of people. But you’re missing the point, of course.”

    Fine. I used OpenLDAP. I’ve used AD and it’s the pits. Slow and halting. Munich used GOsa with LDAP to manage thousands of PCs. RedHat advertises they can manage 10K systems as easily as 1. So, what’s your point? I can manage 100 PCs as easily as 1 just using a few scripts and OpenSSH. I prefer thin clients so I have a further multiplier of my efforts by needing only to tweak a few servers instead of many clients.

  20. TEG wrote, “StatCounter states rather clearly that it uses page views rather than unique IP, so NAT is simply an non-issue for its global stats.”

    They may be clear about that but what’s also clear is that they don’t see any “bump” in the vicinity of rather large migrations. How are those machines hiding? It’s probably something related to the router: NAT, firewall etc. anything that blocks access to sites or allows Statcounter to ignore page-views. They state they are unbiased yet they remove “bots” without any mention of how they do that. Maybe 100 clicks for the weather from one IP address counts as a bot. You’d think that would be OS-neutral but it’s not because we know most retail shelves push that other OS and consumers are much less likely to NAT.

  21. DrLoser says:

    Now, now, Robert. Don’t go getting all excited about this.

    First question here: What do you mean by a “government” distro?

    Federal? Local?

    I’m sure you have an intimate knowledge of how German government works.

  22. Kevin Sorbo says:

    “Schools are medium-sized businesses with a need for effective IT”

    That’s a meaningless blanket statement. It depends on the size of the school. Some deployments are quite small. How many systems in your school up in the Innukistan outback, for instance?

    “Universities are often big businesses.”

    I would classify Universities as large enterprises. You’d blow a gasket if you realized how popular AD, Oracle and Peoplesoft are for internal infrastructure in your average, large university, or how much server-side Java is used. These are practically all heterogenous environments for various reasons.

    But thanks for re-enforcing the point, your experience up in Innukistan or in setting up a thin client at home isn’t quite applicable to a university environment.

    “Who cares how a “corporate environment operates”.”

    Anyone working in one, or a comparably sized deployment should, GPO by itself increases operating efficiency by leaps and bounds. Even in a heterogenous environment, AD and GPO for the windows systems, Puppet or Chef for the *nix systems. You learn a thing or three on efficiency when you’re managing thousands of systems and tens of thousands of people. But you’re missing the point, of course.

    On any sizeable environment, you test patches in a sandbox first, then you deploy them centrally, and you do so during off hours.

    ” Many large corporations see great advantage in GNU/Linux”

    Agreeably so, but that’s not the point either, now is it? So too, do many large corporations see great advantage in Microsoft’s stack, or in proprietary Unix (see Microsoft, HP, Oracle and IBM as their customers), but that’s not really a useful argument now is it?

  23. That Exploit Guy says:

    It’s still true that because that other OS clogs retail shelves it gets most of the play with consumers and those in business/technology who know they have choices usually put up a firewall and NAT things so StatCounter may not count them properly.

    StatCounter states rather clearly that it uses page views rather than unique IP, so NAT is simply an non-issue for its global stats.
    Net Apps, on the other hand, is rather vague when it comes to defining “unique visitors”, though rumour has it that it uses primarily cookies and falls back to IP addresses only when the former is not an option.
    So, no, I don’t see how your “NAT/firewall” argument is supposed to hold up when it comes to operating systems statistics, but, hey, you “know [your] way around the Debian repository”, so that kind of makes you an expert of everything.
    Pfft…

  24. Kevin Sorbo wrote, ” Poggs conflates how his Innuk school operates a handful of workstations with how a corporate environment operates.”

    Schools are medium-sized businesses with a need for effective IT. Universities are often big businesses. Who cares how a “corporate environment operates” if they are wasting money on M$’s stuff? Many large corporations see great advantage in GNU/Linux: Google, IBM, RedHat and all their customers.

  25. Kevin Sorbo says:

    “The idea is that every IT system must be as secure as possible whatever it takes proportionate to the risks.”

    This can be interpreted in many ways. If we were to interpret it in its most literal sense, Linux and Unix, in favour of OpenVMS. If we interpret it in the more rational way, it would imply making a system as secure as is possible within the abilities of the IT staff and chosen platform within the realm of feasibility. in this sense, not even DOS is ruled out, as long as it is made as secure as possible.

    *nix isn’t inherently secure, nor is NT inherently insecure, some are easier to harden than others,

    “That would imply GNU/Linux or UNIX operating systems should be used everywhere. That cuts out M$ and possibly Apple.”

    OS X is a Unix system.
    Tell me, which is “as secure as possible”, Linux with an incompetent IT staff, or NT with vetted, skilled administrators? If your answer is the former, then I advise you to go out more, both your capability for rational thought as well as the size of your social circle (you being the most rational person you know) are quite suspect.

    “It’s more than operating systems though. Will Java make the cut?”

    Server-side Java is what gets Linux’s foot in the door in corporate, server environments. Yes, it’ll make the cut, something like half (conservative estimate) the corporate universe runs on it.

    Even so, limiting to operating systems defeats the intent of the legislation. A secure OS won’t save your from insecure services and applications, or have we forgotten about Heartbleed as the myriad of privilege escalation vulnerabilities?

    Things like RBAC and ACLs help in this regard, but are by no means foolproof.

    “That could cut out a lot of non-Free software right there.”

    That would also cut out a lot of free software, too.

    “How can you certify binary-only updates made at noon on a work-day?”

    Because domains, sandboxes and IT staffs don’t exist. Every workstation has automatic updates on and updates itself individually (pro-tip: in any sane environment of large scale, patches are tested first, and pushed to workstations from the domain controller, usually during off hours).

    Once again, Poggs conflates how his Innuk school operates a handful of workstations with how a corporate environment operates.

  26. DrLoser says:

    Nonsense. I am the most rational person I know.

    Slippery slope, Robert, slippery slope.

    For example, I am the most hysterically funny person I know. Obviously I have tested this theory. I’ve done so over and over again. Nobody makes me laugh as much as I make other people laugh.

    I’m a laugh riot! (Also, I’ve got a Silver Medal in being rational. It’s a non-profit sideline.)

    Strangely, Robert … Strangely … Nobody else believes me.

    Well, apart from that weirdo drug-addict down the street. But then again, she’ll laugh at anything. Maybe even nothing.

    Is there the slightest possibility that you are overestimating yourself here?

    You know, what with never once taking other peoples’ observations on the matter into account?

  27. DrLoser wrote of me and “my kind”, “Reasoning about obvious statistics doesn’t come naturally”.

    Nonsense. I am the most rational person I know and I do care about statistics which is why I use StatCounter, such as it is. If I didn’t care, I would use IDC’s 3% number from way back in time and extrapolate with ~30% per annum growth. It’s still true that because that other OS clogs retail shelves it gets most of the play with consumers and those in business/technology who know they have choices usually put up a firewall and NAT things so StatCounter may not count them properly. Same with NetApplications. Remember Silicon Valley? Google switched to GNU/Linux and really affected their stats far more than Google’s numbers would warrant. NetApplications’ number for GNU/Linux nearly trebled if you count California compared to omitting it. Too bad they deleted that freebie. They won’t let me see Ethiopia and my other favourites for $0 any longer.
    “Google, a mere 10K people in a small region of California, changes the stats markedly throughout USA, from 0.65% excluding California to 1.72% including California. It is no wonder that NetApplications shows too high values for that other OS when business-usage is weighted so heavily.”

  28. DrLoser says:

    The insidious effect of retail shelf-space again. Perhaps the GNU/Linux-share is 1% geeks and 1.5% really ticked-off consumers.

    Ah, yes, that insidious effect again, Robert. I’m surprised that Sir Isaac Newton didn’t include it as his Fourth Principle.

    I think a nine-month jump from 3.1% to 11.7% requires a slightly more sophisticated retort, don’t you? Take your pick:

    “Microsoft are just cannibalising their own base.” (Surprised you didn’t pick that one.)
    “That’s nothing like as stable as the German Linux growth, assuming a normal distribution. Watch out for January 2050!” (Not surprised you didn’t pick that one.)
    “It’s all about Secure Boot and UEFI lock-in!” (Presumably you didn’t pick that one because it was last year’s pathetic excuse.)

    Never mind, Robert. We’ll always have Uruguay. Or maybe it’s Ethiopia? Whichever.

    Reasoning about obvious statistics doesn’t come naturally to those of us who have an incorrigible bias, does it?

  29. olderman wrote, “There is nothing enjoyable about having to tinker to get a job done, or make due when the software you want doesn’t exist or is inferior crap.”

    Tinker? Sure I do but that’s just me trying to perfect the PC/server. The ordinary joe is quite satisfied with GNU/Linux and the most commonly used applications.

    Ken Starks has donated GNU/Linux PCs to a lot of kids in Texas and several of them have gone on to graduate university. As he writes: “Currently, five of our former kids are in graduate school, everyone of them studying engineering, physics or computer sciences Did the computer we gave them as a kids have anything to do with where these students are today? I can’t prove it did any more than you can prove it didn’t.” GNU/Linux works for people. I’ve seen a lot of students helped by GNU/Linux. First thing, I could give more computers that worked to students using GNU/Linux. That’s a huge advantage of FLOSS. In all the years we used GNU/Linux in schools, very few complained the platform or the applications were second-rate. Most were amazed that there was an alternative and that it worked so well. Organizations large and small would not be migrating to GNU/Linux if GNU/Linux and FLOSS didn’t work for them.

  30. DrLoser wrote, “The despised and despicable Windows 8.1 has moved from 3.1% in December 2013 … comfortably outscoring Linux at a percentage of 2.7% … to a whopping 11.8% in September 2014!”

    The insidious effect of retail shelf-space again. Perhaps the GNU/Linux-share is 1% geeks and 1.5% really ticked-off consumers.

  31. DrLoser says:

    Well, obviously, it’s all about UEFI lock-in.

  32. DrLoser says:

    That’s a Hell of a lot better than ~1%, eh? Average for the last two weeks in August was 2.56%.

    Out of mild interest, Robert, I pulled down the figures from StatCounter in Germany since Jan 1st 2013. No particular date: just long enough to show possible trends.

    In Germany, with an acknowledged Linux base, the thing has flat-lined for eighteen months or more. Excel gives me an average of 2.55%. It did very well in August and September last year — over 3%! — should you wish to find a correlation of some sort. The baseline seems to be something like 2.2% at the moment.

    Also, bad news from Germany in the same period! The despised and despicable Windows 8.1 has moved from 3.1% in December 2013 … comfortably outscoring Linux at a percentage of 2.7% … to a whopping 11.8% in September 2014!

    These Huns. They just don’t “grok” Freedom the way you and I do, nicht wahr?

  33. olderman says:

    “The folks using GNU/Linux are enjoying it and getting great IT for a reasonable price. ”

    Baloney. There is nothing enjoyable about having to tinker to get a job done, or make due when the software you want doesn’t exist or is inferior crap.

  34. wolfgang wrote, “deutschland?”

    2.5% for August. That’s a Hell of a lot better than ~1%, eh? Average for the last two weeks in August was 2.56%. Finland, Italy, Germany, Spain and France are all similar. That’s a good chunk of the continent.

    Back in July, wolfgang wrote, “linux means is dragging along at 1% or so. “

    Nope. The global average is up to ~1.4%. It just keeps growing. I’ve been looking into planting more trees in my yard. Did you know oaks start small but become huge and live hundreds of years? GNU/Linux is like that. There’s just no reason for it to die and every reason to continue growing. Unlike some non-Free stuff, a low rate of growth does not mean bankruptcy, just more fun. The folks using GNU/Linux are enjoying it and getting great IT for a reasonable price. There’s no rule on how many folks have to use it but it’s good so the numbers keep growing.

  35. wolfgang says:

    …all over…

    pogson such big fan of statcounter in uruguay so why not look up how famous linux in
    deutschland? see linux on chart, but behind windows 7, windows 8, windows 8.1, windows xp, windows vista, and mac. deutsch menschen not getting on board at local station, es scheint!

Leave a Reply